Instructions to use ypwhere/LQK-BabyLM-Strict-Small with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use ypwhere/LQK-BabyLM-Strict-Small with Transformers:

# Use a pipeline as a high-level helper
from transformers import pipeline

pipe = pipeline("text-generation", model="ypwhere/LQK-BabyLM-Strict-Small", trust_remote_code=True)

# Load model directly
from transformers import AutoModelForCausalLM
model = AutoModelForCausalLM.from_pretrained("ypwhere/LQK-BabyLM-Strict-Small", trust_remote_code=True, device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use ypwhere/LQK-BabyLM-Strict-Small with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "ypwhere/LQK-BabyLM-Strict-Small"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "ypwhere/LQK-BabyLM-Strict-Small",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/ypwhere/LQK-BabyLM-Strict-Small

SGLang

How to use ypwhere/LQK-BabyLM-Strict-Small with SGLang:

Install from pip and serve model

# Install SGLang from pip:
pip install sglang
# Start the SGLang server:
python3 -m sglang.launch_server \
    --model-path "ypwhere/LQK-BabyLM-Strict-Small" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "ypwhere/LQK-BabyLM-Strict-Small",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker images

docker run --gpus all \
    --shm-size 32g \
    -p 30000:30000 \
    -v ~/.cache/huggingface:/root/.cache/huggingface \
    --env "HF_TOKEN=<secret>" \
    --ipc=host \
    lmsysorg/sglang:latest \
    python3 -m sglang.launch_server \
        --model-path "ypwhere/LQK-BabyLM-Strict-Small" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "ypwhere/LQK-BabyLM-Strict-Small",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use ypwhere/LQK-BabyLM-Strict-Small with Docker Model Runner:
```
docker model run hf.co/ypwhere/LQK-BabyLM-Strict-Small
```

LQK-BabyLM-Strict-Small / config.json

ypwhere

Publish LQK strict-small final model

94b1d44 verified about 1 month ago

Raw

History Blame Contribute Delete

1.55 kB

	{
	"architectures": [
	"NanoGPTForCausalLM"
	],
	"attn_res_block_size": 8,
	"auto_map": {
	"AutoConfig": "modeling_nanogpt.NanoGPTConfig",
	"AutoModel": "modeling_nanogpt.NanoGPTModel",
	"AutoModelForCausalLM": "modeling_nanogpt.NanoGPTForCausalLM",
	"AutoModelForMaskedLM": "modeling_nanogpt.NanoGPTForMaskedLM"
	},
	"bias": false,
	"bidirectional": false,
	"block_size": 512,
	"bos_token_id": 0,
	"dropout": 0.0,
	"eos_token_id": 0,
	"layer_norm_epsilon": 1e-05,
	"model_type": "nanogpt",
	"n_embd": 512,
	"n_head": 8,
	"n_layer": 32,
	"nanogpt_checkpoint": {
	"filename": "ckpt_final-w0090M-i000471.pt",
	"git_sha": "11a6860211f85a96fcb0cfd30a0e763eda931d2a",
	"iter_num": 471,
	"labels": [
	{
	"actual": 90343884,
	"name": "words_final",
	"revision": "chck_90M",
	"series": "words",
	"target": 90343884
	},
	{
	"actual": 123469824,
	"name": "tokens_final",
	"revision": "chck_123M",
	"series": "tokens",
	"target": 123469824
	}
	],
	"num_updates": 471,
	"role": "final",
	"sha256": "5e1e936ce311c6535526e52d1a97059b587010c9b4e2dc3e5ee1b94fa742d512",
	"tokens_seen": 123469824,
	"words_seen": 90343884
	},
	"pad_token_id": 0,
	"rope_theta": 10000.0,
	"swiglu_mult": 2.6666666666666665,
	"torch_dtype": "float16",
	"transformers_version": "4.51.3",
	"use_attn_gate": true,
	"use_attn_res": true,
	"use_rmsnorm": true,
	"use_rope": true,
	"use_swiglu": true,
	"vocab_size": 16000
	}